Geographia Technica, Vol 21, Issue 2, 2026, pp. 182-195
SEASONAL REORGANIZATION OF CLIMATE SIMILARITY IN THE CARPATHIAN BASIN: A TIME-BASED CLIMATE NETWORK APPROACH
Attila MAGYARI-SÁSKA 
ABSTRACT: This study examines changes in the seasonal organisation of climate in the Carpathian Basin using a time-based climate-network approach. In contrast to conventional climate networks, where nodes usually represent spatial locations, the present method defines individual year–month observations as network nodes. Edges describe multidimensional climatic similarity based on monthly values derived from daily temperature, precipitation, humidity and radiation data. The analysis uses E-OBS gridded observational data for ten selected locations and compares two 30-year periods: 1961–1990 and 1995–2024. The results show that the annual climatic cycle remained broadly recognisable in the recent period, with stable winter and summer components still present in the network structure. However, the boundaries between seasonal groups became less sharply defined. The strongest changes were associated with transition months, especially February, March, April, September and October. Shannon entropy indicates that some of these months became more fragmented, while prototype similarity matrices reveal stronger cross-seasonal relationships between spring and autumn states. Percolation analysis shows only minor changes in overall network robustness, suggesting that the observed transformation should not be interpreted as a collapse of the seasonal structure, but rather as a moderate reorganisation of the relationships between monthly climatic states. The study demonstrates that time-based climate networks can complement conventional trend-based climate analyses by highlighting structural changes in the temporal organisation of seasonality.
Keywords: time-based climate networks; seasonal reorganisation; Carpathian Basin; climatic similarity; Shannon entropy; percolation analysis
Full article here